Agentic Payments Statistics 2026: Market Size, Adoption, and Growth Data

In April 2026, Juniper Research put total agentic commerce transaction value at roughly $8 billion for 2026, with a forecast of $1.5 trillion by 2030.

Growth Centr publishes evergreen, research-backed analysis for founders, operators, and payments teams who need the actual numbers rather than the narrative around them.

Agentic payments sit at an awkward stage right now: the infrastructure announcements are enormous, the forecasts are enormous, and the live transaction volumes are still small enough to count.

This report brings together the market size estimates, adoption rates, protocol data, conversion benchmarks, and risk statistics published between late 2025 and August 2026, so you can see where the signal ends and the projection begins.

Key Takeaways

• Agentic commerce transaction value sits near $8 billion in 2026, per Juniper Research.
• Only 3% of merchant transactions currently involve an AI agent.
• Gartner projects $15 trillion in agent-intermediated B2B purchases by 2028.
• AI-referred retail traffic converted 42% better than non-AI traffic in March 2026.
• Trust and liability, not technology, remain the primary constraints on adoption.

What Counts as an Agentic Payment

Definitions drive almost every discrepancy in the figures below, so it is worth separating the categories before reading the numbers.

Visa's joint research with Artemis splits the space into two behaviours. Macro commerce covers an agent acting for a person, booking a flight or managing a subscription, at consumer-sized values that sit comfortably on existing card rails.

Micro commerce covers small, frequent payments made by one piece of software to another, usually for an API call or a slice of compute, and these typically come in well under a dollar.

That distinction matters because most published "agentic commerce" market sizing measures retail baskets, while the fastest-growing live volume by transaction count is machine-to-machine micropayments.

A third category, agent-influenced commerce, counts purchases where an AI assistant shaped the decision but a human completed checkout. McKinsey's widely cited trillion-dollar figures use this broader definition, which is why they run orders of magnitude above measured transaction data.

Agentic Payments Market Size in 2026

The 2026 baseline is small. Juniper describes 2025 and 2026 as pilot years, and its study concluded that trust remains the number one barrier and that agentic commerce will complement rather than replace traditional eCommerce checkouts for the foreseeable future.

Surrounding market estimates vary by scope:

Grand View Research valued the global agentic commerce market at $5.7 billion in 2025 and projected $7.7 billion in 2026, reaching $65.5 billion by 2033 at a 35.7% CAGR, with North America holding 38.2% of 2025 revenue.

Note that this measures platform and enabling technology revenue, not transaction value, which is why it lands near Juniper's transaction figure by coincidence rather than agreement.

Growth Forecasts Through 2030 and Beyond

Forecasts diverge by roughly 25x depending on definition and geography.

Juniper's $1.5 trillion projection sits between Bain's US estimate of $300 billion to $500 billion by 2030 and Morgan Stanley's $190 billion to $385 billion range for US e-commerce, while McKinsey's $3 trillion to $5 trillion global figure sets the ceiling.

The spread is not a sign that one firm is wrong. It reflects whether a forecast counts autonomous checkout only, or every purchase an agent touched. When you read a headline number, check which one it is.

Consumer Adoption Statistics

Consumer usage of AI in the buying journey is now mainstream. Delegating the payment itself is not.

• Visa research found 47% of US shoppers use AI tools for at least one shopping task, from price comparisons to recommendations.
• IBM's Institute for Business Value reported in January 2026 that 45% of consumers already use AI for some part of the buying journey, and ChatGPT handles roughly 50 million shopping queries per day.
• Adyen found UK shopper use of AI assistants doubled from 12% to 28% in a year, and US usage more than doubled from 12% to 35%, with 51% of US shoppers willing to let AI handle the entire process including the final purchase.
• Checkout.com's June 2026 report found 33% of consumers expect at least 10% of their purchases to be AI-driven within a year, with interest at 67% in the US, 62% in Australia and around 58% in the UK.

The counterweight is sharp. In the same Checkout.com dataset, 24% of consumers said they will never delegate purchases to AI and 27% trust no organisation to operate an agent.

A YouGov survey of 2,080 UK adults for ACI Worldwide in June 2026 found just 19% trust AI agents to follow guidance on day-to-day purchasing, against 55% who trust a human expert.

The pattern repeats across markets. A 2026 Logica Research survey for Commerce and PayPal found 63% of Australian shoppers interested in trying agentic AI shopping tools, but only 4% wanted AI assistance at the point of payment.

What Consumers Require Before Delegating Spend

Checkout.com surveyed 12,005 consumers across the UK, US, Brazil, China, France and the UAE, plus 400 heads of payment at consumer-facing merchants in the UK and US. The control conditions were consistent:

• Spending caps: 30%
• Instant revocation: 29%
• Easy cancellation: 28%
• Average per-purchase spend consumers would authorise without additional approval: £177, below the £200 merchants assumed digiral transactions.

That £177 ceiling is one of the more useful operational numbers in the entire dataset. It tells you which basket sizes are addressable today and which are not.

For teams designing mandate structures around these limits, our breakdown of the AP2 Agent Payments Protocol covers how signed spending mandates are actually constructed.

Merchant Readiness and the Execution Gap

Merchant intent runs far ahead of merchant activity.

• 89% of merchants are actively preparing for agentic commerce, yet only 3% of transactions currently involve an AI agent.
• 72% of merchants agree consumers will adopt agent-led shopping faster than most merchants are prepared for.
• 75% of merchants say real-time permission revocation will be critical to adoption.
• The Merchant Risk Council found roughly 63% of merchants exploring or planning agentic AI payment implementation, with 72% already using payment tokenization and 43% accepting real-time payments.

The 89% versus 3% gap is the single most quoted statistic in the category for good reason. It quantifies a market where nearly everyone is building and almost nothing is live.

The OpenAI Instant Checkout Case Study

The clearest correction to 2026 hype came from the most visible product in the space.

OpenAI launched Instant Checkout inside ChatGPT in September 2025 and walked it back by early 2026, telling Modern Retail that Instant Checkout was moving to Apps where purchases can happen more seamlessly.

By February 2026, roughly 30 Shopify merchants were live on Instant Checkout, according to Forrester principal analyst Emily Pfeiffer, against the million-plus merchants once promised.

The conversion data explains why. Walmart made around 200,000 products available through Instant Checkout from November 2025, and EVP Daniel Danker reported that checkout completed inside ChatGPT converted at about one-third the rate of sending the same shoppers to walmart.com.

Only 17% of surveyed marketplace shoppers across five markets said they felt comfortable completing purchases with AI, according to ChannelEngine's Marketplace Shopping Behavior 2026 report.

The protocol survived the product. For the commercial detail on what it costs to participate, see our guide to ChatGPT Instant Checkout fees and requirements.

Protocol and Infrastructure Adoption Data

Five specifications now compete for the authorisation layer, and merchants are increasingly supporting more than one.

• ACP (Agentic Commerce Protocol): built by OpenAI and Stripe, live since September 2025, reaching over a million Stripe-connected merchants by design.
• UCP (Universal Commerce Protocol): announced by Google at NRF in January 2026 with Etsy, Shopify, Target, Wayfair and Walmart, endorsed by Adyen, American Express, Mastercard, PayPal, Stripe, Visa and Worldpay, and compatible with A2A, AP2 and MCP.
• Visa Intelligent Commerce and Trusted Agent Protocol: over 100 partners globally, more than 30 building in the VIC sandbox, and over 20 agents or agent enablers integrating directly.
• Mastercard Agent Pay: agentic tokens plus Verifiable Intent, extended in June 2026 with a machine-to-machine variant supporting sub-cent payments.
• AP2 (Agent Payments Protocol): Google's vendor-neutral mandate specification, now with stablecoin extensions.

Cross-protocol support appears to pay. Merchants supporting multiple protocols see roughly 40% more agentic traffic. The underlying plumbing draws on the Model Context Protocol for tool access and the A2A agent-to-agent protocol for inter-agent messaging.

Machine-Native Micropayments: The Onchain Numbers

This is where measured volume, rather than forecast volume, actually exists.

x402, incubated by Coinbase and Cloudflare and now stewarded by the Linux Foundation, has processed roughly $15.0 million in adjusted volume across 109.6 million transactions since launching in May 2025, with most activity on Base, Solana and Polygon.

The Machine Payments Protocol, built by Stripe and Tempo with Visa contributions, settled about $25,000 across roughly 115,000 transactions in its first weeks after a mid-March 2026 launch.

The average payment on both is a fraction of a cent. A fixed card fee on a transaction that small would cost far more than the payment itself, which is why newer settlement rails matter here: they have pushed costs down enough to make payments in the one-cent to one-dollar range practical for the first time.

Read those two figures carefully. 109.6 million transactions is real machine-native demand. $15 million in value is not a market.

Both facts are true at once, and our comparison of stablecoins versus card rails for AI agent payments works through which flows belong on which rail.

B2B Agentic Payments Statistics

The B2B numbers are larger than the consumer numbers and move on a shorter timeline.

• Gartner forecasts that AI agents will intermediate more than $15 trillion in B2B purchases by 2028, with a definition requiring meaningful autonomous action across supplier identification, contract negotiation and order execution.
• The same prediction holds that 90% of all B2B buying will be AI agent intermediated by 2028.
• Forrester's 2026 State of Business Buying, covering around 18,000 global B2B buyers, found 94% use generative AI as a core research tool, and Gartner's March 2026 survey of 646 B2B buyers found 67% prefer a rep-free experience for routine procurement.
• Gartner separately projects that 20% of monetary transactions will be programmable by 2030.

Execution risk is material. Gartner expects more than 40% of agentic AI projects to be cancelled by the end of 2027, driven by escalating costs, unclear business value and inadequate risk controls, while only 21% of organizations report a mature governance model for autonomous agents and 52% cite data quality as the biggest deployment blocker.

Finance teams evaluating this shift will find the operational view in our piece on autonomous agents in B2B finance and accounting workflows.

Traffic, Conversion, and Revenue Performance

Agent-influenced traffic is now a measurable, high-quality channel even where agents do not complete the payment.

• AI-driven traffic to US retail sites grew 393% year over year in Q1 2026, following a 693% year-over-year rise during the November to December 2025 holiday period, per Adobe Analytics.
• In March 2026, AI referral traffic converted 42% better than non-AI traffic, reversing a 38% deficit twelve months earlier, and produced 37% more revenue per visit with 48% longer time on site.
• AI-referred traffic grew 138% year over year in May 2026 and 1,324% since October 2024, based on more than 1 trillion visits to US retail sites.
• Shopify reported AI-referred orders growing nearly 13-fold in Q1 2026, with conversion rates almost 50% higher and average order values 14% higher.
• Salesforce estimated AI agents influenced 20% of global retail orders during the 2025 holiday season, worth roughly $262 billion.

Attribution remains the weak point. As much as 70% of AI referral traffic is misclassified as direct because AI tools often do not pass referrer headers, and Google Analytics only added a native AI Assistant channel in May 2026.

If your dashboard shows negligible AI traffic, measure again before concluding anything. Our guide to AI search visibility for B2B SaaS covers the tracking side of this.

Fraud, Risk, and False Decline Data

Agentic transactions break the assumptions fraud models were built on.

• Most agent-initiated transactions would be declined today, because there is no reliable standardised way to distinguish a permissioned agent from a malicious automated attack.
• Visa's PERC team reported a more than 450% increase in dark web posts referencing AI agents over a six-month period.
• False declines cost merchants roughly 13 times more than actual fraud according to Javelin-attributed studies, yet only around 64% of merchants track their false decline rate, and just 32.4% use 3D Secure 2.
• First-party fraud now represents around 36% of reported fraud, up from roughly 15% in 2023.

The structural problem is liability. Chargeback windows and evidence rules were built for human-speed commerce with one clear order to point to, and once agents transact thousands of times an hour through chains of other agents, there is no settled way to unwind a payment that went wrong.

Our breakdown of agentic payments security risks goes deeper on the specific attack surfaces.

How to Read These Numbers

Three cautions apply to every figure above.

First, methodology varies wildly. A $5 trillion forecast counting AI-influenced purchases and a $8 billion figure counting completed agentic transactions are not comparable, even though both get published under the same headline term.

Second, most adoption data is survey-based intent, not observed behaviour. The gap between the 67% of US consumers expressing interest and the 3% of transactions involving an agent is the gap between stated and revealed preference.

Third, the vendors publishing the data mostly sell into the category. Visa, Mastercard, Stripe, Checkout.com and Adobe all have commercial exposure to the outcome. That does not make the numbers wrong, but it does make directional treatment sensible.

Conclusion

Growth Centr exists to give operators the reference numbers behind fast-moving categories, updated when the evidence changes rather than when the news cycle turns.

The 2026 picture on agentic payments is consistent across sources: measured transaction volume is still in the single-digit billions, infrastructure commitment from card networks and platforms is close to universal, consumer interest is high while consumer trust is low, and the largest projected value sits in B2B procurement rather than retail baskets.

The protocols have largely stopped competing and started converging. What has not been solved is liability, dispute handling, and agent identity, and those three problems, more than model capability, will determine how quickly the forecasts above turn into settlement volume.

Read Next

• Stablecoins vs Card Rails for AI Agent Payments: Which Settlement Layer to Choose
• ChatGPT Instant Checkout Fees and Requirements: What Merchants Need to Know
• AI SDRs for B2B SaaS: Where They Work and Where They Break (2026)

FAQs

1. What is the agentic payments market size in 2026?

The agentic payments market size in 2026 is approximately $8 billion in total agentic commerce transaction value according to Juniper Research, with Grand View Research separately valuing the agentic commerce technology market at $7.7 billion for the same year. Both figures reflect a pilot-stage market rather than mainstream deployment.

2. What percentage of transactions involve an AI agent in 2026?

The percentage of transactions that involve an AI agent in 2026 is roughly 3%, according to Checkout.com research published in June 2026, even though 89% of surveyed merchants report actively preparing for agentic commerce.

3. How fast is agentic commerce expected to grow through 2030?

Agentic commerce is expected to grow through 2030 from roughly $8 billion in 2026 to $1.5 trillion globally, per Juniper Research, with competing estimates ranging from Morgan Stanley's $190 billion US base case to McKinsey's $3 trillion to $5 trillion global range depending on whether AI-influenced purchases are counted.

4. How much will consumers let an AI agent spend without approval?

Consumers will let an AI agent spend an average of £177 per purchase without additional approval, based on Checkout.com's survey of 12,005 consumers across six markets, which is below the £200 that UK and US merchants assumed.

5. Which agentic payment protocols have the most adoption in 2026?

The agentic payment protocols with the most adoption in 2026 are ACP from OpenAI and Stripe, Google's UCP, Visa Intelligent Commerce with Trusted Agent Protocol, Mastercard Agent Pay, and AP2, with x402 leading measured machine-native transaction counts at 109.6 million transactions since May 2025.

Disclaimer:
This content is provided for informational purposes only and does not constitute legal, financial, or compliance advice. Protocol versions, governance arrangements, and partner counts cited here reflect publicly announced milestones as of August 2026 and are moving quickly. Adoption figures come from vendor and foundation announcements with differing methodologies and should be treated as directional signals rather than guaranteed outcomes.